2026-10-02T22:17:51Z https://dcpapers.dublincore.org/oai

oai:dcpapers.dublincore.org:952626237 2026-09-10 dcmi-2026 openaire

Human-Centered TEI Encoding as Infrastructure for Meaning-Driven AI in Portuguese Literature Giménez Celano, Diego Emanuel This paper argues that the alignment of artificial intelligence with human values begins not at the level of algorithms but at the level of knowledge modeling. Drawing on the ongoing TEI encoding of Os Lusíadas by Luís de Camões, it introduces metadata as a form of epistemic governance that structures what AI systems can recognize, process, and reproduce as knowledge. Rather than positioning AI as an interpretative agent, the study frames it as operating within a semantic architecture designed by human editors. TEI encoding is therefore approached not merely as technical markup but as an interpretative infrastructure in which scholarly values, transparency, plurality, and cultural responsibility are formalized into structured data. The paper proposes a human-centered workflow in which AI assists editorial practices through schema-constrained suggestions, consistency checks, and semantic pattern detection while preserving human authority over interpretative decisions. Attention is given to the ethical implications of encoding a canonical early modern epic whose textual tradition intersects with mythology, religion, and historical memory. By conceptualizing metadata as a mechanism of epistemic governance, this study shifts the discussion of meaning-driven AI upstream and argues that responsible machine participation in cultural knowledge depends primarily on the design of metadata environments. Dublin Core Metadata Initiative 2026-09-10 info:eu-repo/semantics/conferenceObject Text application/pdf text/html https://doi.org/10.23106/dcmi.952626237 https://dcpapers.dublincore.org/article/952626237 Dublin Core Metadata Initiative Conference Proceedings eng info:eu-repo/semantics/openAccess https://creativecommons.org/licenses/by/4.0/